Page 1
International Journal of Automotive and Mechanical Engineering (IJAME)
ISSN: 2229-8649 (Print); ISSN: 2180-1606 (Online);
Volume 13, Issue 3 pp. 3728 - 3741, December 2016
©Universiti Malaysia Pahang Publishing
DOI: https://doi.org/10.15282/ijame.13.3.2016.15.0305
3728
Propagation and scattering of guided waves in composite plates with defects
Bibi I. S. Murat1* and P. Fromme2
1Faculty of Mechanical Engineering, Universiti Teknologi MARA,
40450 Shah Alam, Selangor, Malaysia *Email: [email protected]
Phone: +60355435224; Fax: +60355435160 2Department of Mechanical Engineering, University College London,
WC1E 7JE London, UK
ABSTRACT
Failure in aerospace composites owing to low-velocity impact raises a significant
maintenance concern because it can lead to invisible damage. For the aerospace industry,
such defects pose a potential danger to the structural integrity of aircraft. This in turn
jeopardises passenger safety and incurs high repair costs. Hence, it is important to
efficiently monitor composite structures during the service life. In this study, the potential
of low-frequency guided ultrasonic waves for health monitoring in laminated composite
plates is investigated. This study focuses on the use of the first antisymmetric guided
wave mode (A0). The first part of this study is to investigate the propagation of the
A0 mode in three different undamaged composite plates experimentally. The dispersive
and anisotropic behaviour are in agreement with the results of finite element simulations
and semi-analytical analysis. The final part of this study presents the scattering of guided
waves at the impact damage using a non-contact laser interferometer. Significant
scattering activities were observed and the impact damage size can be estimated to be
about 10 × 25 mm. In conclusion, these results demonstrate the potential of guided
ultrasonic waves for the inspection of aerospace composite structures.
Keywords: Composite plates, guided ultrasonic waves, non-destructive testing.
INTRODUCTION
In general, aerospace composites, in the form of carbon fibre laminates, consist of layers
of polymer matrix reinforced with high-strength carbon fibres. In a complete investigation
by Richardson and Wisheart [1], it is stated that one major concern related to composite
laminates is their susceptibility to sustain low-velocity impact damage. The problem with
low-velocity impact damage in composites is that it is often not visible or is barely visible
in a typical visual inspection [2]. Shyr et al. [3] found that visible damage can be clearly
detected and remedial action could be taken immediately to maintain the structural
integrity. However, this is often not the case for impact damage in composites. A major
concern is the growth of hidden, undetected defects caused by low-velocity impacts and
fatigue [4]. Various different failure modes and mixed damage modes may occur [2, 5].
Matrix cracking, delamination, fibre debonding and fibre breakage are examples of
various failure modes under low-velocity impact [6, 7]. Wisnom [8] highlights that failure
to detect these internal damages at an early stage may result in a catastrophic failure of
the composite structure. In order to maintain the quality and reliability of a composite
structure, non-destructive testing (NDT) is commonly used. Visual inspection, ultrasonic
Page 2
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3729
testing [9], acoustic emission [10], X-ray radiography [11] and eddy-currents [12] are
amongst the NDE methods employed in aerospace inspection. However, for the large
aircraft structures most methods are very time consuming and costly, and interrupt the
aircraft service. This indicates a need for rapid inspection and cost-effective methods for
monitoring large composite structures.
One possible method, the guided ultrasonic waves NDE method, has been chosen
to be further explored in this study [13]. Using low excitation frequency, guided waves
can propagate over long distances with limited energy loss. From a single location, the
guided waves can cover large areas, which helps to reduce the inspection time. The
reflection of the propagating wave at defects enables rapid detection of defects in large
structures [14]. This method has been used successfully for the detection of defects in
large metal plates and long pipes, i.e., for corrosion and crack detection [15]. However,
the behaviour of the guided waves is somewhat more complicated in composite structures
owing to the physical properties of composites that are generally inhomogeneous and
anisotropic in nature. The capability of the guided waves for the inspection of aerospace
structures is still under investigation. Many factors could affect the wave propagation and
scattering [16-18]. The properties of guided waves in anisotropic plates are more
complicated than those in isotropic plates [19, 20]. The plate geometry, material
properties, fibre arrangement, fibre orientation, transducer frequency, excitation mode
and type of impact damage are among the factors. Moreover, multiple reflections can
form an infinite number of wave modes through the thickness. Wilcox et al. [21] identified
that the modes can be either symmetric, noted as Sn (S0, S1, S2.....Sn), or antisymmetric,
noted as An (A0, A1, A2…..Sn), and these modes are generally dispersive [22]. Each wave
mode has a different speed, a different wavelength and a different wave pattern (mode
shape) across the thickness, which can add to the complexity of the received signals [22].
For waves propagating in composite laminates, the wave interaction depends on many
factors such, as the excitation frequency, the geometry of the structure, material
properties, direction of propagation and interlaminar conditions [23-25]. Although the
benefits of using guided waves are huge, these factors describe the difficulties in using
guided waves for composite inspection. Hence, knowledge of the properties of guided
wave propagation in composites is important for the successful implementation in non-
destructive evaluation. Therefore, the objective of this study is to investigate the potential
of guided ultrasonic waves for detecting impact damage in composite plates. This study
aims to achieve a better understanding of guided wave propagation in composite plates
and their interactions with impact damage. The outcomes of this research will help to
establish an efficient technique for the inspection of composite materials using guided
ultrasonic waves.
METHODS AND MATERIALS
Two sets of guided wave experiments were performed. The first set of experiments was
performed on four undamaged composite plates with different thicknesses and material
properties. The second experiment was performed on two defective composite plates for
the detection and characterisation of impact damage. Details can be found in Table 1.
The first plate was a large cross-ply carbon fibre plate. The plate was constructed
from 24 prepreg plies in alternating [0°/90°] orientations with symmetry at the mid-plane.
The material of the prepregs was HEC fibre (60%)/SE84 HT epoxy (40%). Each ply has
a nominal thickness of 0.15 mm, giving a total plate thickness of 3.6 mm. The second
plate was a small unidirectional plate. The plate was constructed from 24 prepreg plies in
Page 3
Propagation and scattering of guided waves in composite plates with defects
3730
parallel alignment [0°]. The material and total thickness were the same as those for the
first plate. The final two plates were small cross-ply plates, consisting of eight prepreg
layers with a symmetric layup sequence of [0/90]. The plates were manufactured using
carbon-fibre Tenax HTS (65%) pre-impregnated with Cytec 977-2 epoxy resin (35%).
The ply thickness was 0.25 mm and eight plies were used, giving a plate thickness of 2
mm. The centres of the plates were subjected to a 7.4 J impact using a hemispherical 15
mm impactor head, following standard drop weight impact procedures. The size of the
impact damage can be estimated to be about 10 mm in length and 20 mm in width.
Table 1. Details of the composite specimens.
Test plates Large cross-ply UD plate Small cross-ply
Materials HEC Fiber (60%) / SE84
HT epoxy (40%)
HEC Fiber (60%) /
SE84 HT epoxy (40%)
Carbon Tenax HTS (65%) /
Cytect 077-2 epoxy resin
(35%)
No. of plies 24 24 8
Orientation of
plies
[0/90/0/90/0/90/0/90/0/90/
Symmetry at mid-plane]
All plies in 00 direction [0/90/0/90/ Symmetry at
mid-plane]
Thickness (mm) 3.6 3.6 2
Impact testing (J) - - 7.4
Figure 1. Schematic diagram of the guided ultrasonic wave experimental setup.
Figure 1 shows the setup for the guided wave measurements, which consists of a
modular scanning rig controlled via LabView from a computer. The excitation signal is
generated by a function generator as a voltage signal, amplified by a wide band amplifier
and then applied to a piezoelectric transducer. A laboratory-made piezoelectric
transducer, polarised through the thickness, was used to excite the A0 wave mode. The
discs act in good approximation as a point source and the waves propagate radially
outwards. A brass backing mass was carefully bonded to the piezoelectric disc using
epoxy glue. Electrical voltage is applied via a copper wire that was soldered to the backing
mass. The piezoelectric transducer was directly bonded onto a thin layer of silver coated
paint (Electrolube SCP03B) on the composite plate. The use of silver paint acts as ground
connection for the transducer. When the voltage is applied to the piezoelectric transducer,
the piezoelectric disc contracts and expands. This generates a vertical force to the plate
surface and excites primarily the A0 mode. Good and repeatable signals were obtained.
The displacement field in the specimen is measured using a heterodyne laser vibrometer,
Page 4
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3731
controlled by the scanning rig, which moves parallel to the specimen. The laser
interferometer was moved to perform line scans (in the 0° direction) over a length of 200
mm from the excitation transducer with a 1 mm step size. The time traces (voltage signal)
of the received signals were filtered using a bandpass filter. The voltage signal
corresponds to the velocity of the out-of-plane-displacement of the specimen surface. The
signals were recorded and averaged using a digital storage oscilloscope to improve the
signal-to-noise ratio. The function generator triggers the oscilloscope so that the
excitation and measurement start at the same time. The measured time traces were then
transferred to the computer and further analysed using MATLAB.
For the analysis of wave propagation in undamaged specimens, the phase velocity
(Cp) can be determined by using Eq. (1). Fast Fourier Transform (FFT) was applied to the
signal time traces to obtain the phase change at a given frequency (f.) The phase angle (φ)
difference between two points spaced 1 mm apart was calculated and the values were then
used in the Eq. (1) to calculate the phase velocity. The phase velocities were plotted
against the frequency thickness product (f.d). Meanwhile, the group velocity (Cg) can be
calculated using Eq. (2), where x and t are respectively the distance and arrival time
between two readings. To calculate the group velocity, the arrival time of the signal
envelopes obtained using the Hilbert transform were used to estimate the propagating
time between two measurement points spaced 100 mm apart. The measured phase and
group velocities were then compared to the finite element and semi-analytical predictions.
Meanwhile, the wave attenuation owing to material absorption is assumed to result in
exponential decay of the amplitude with distance of propagation, as defined in Eq. (3),
where A1, A2, r1 and r2 are respectively the amplitudes and radial distances at two
measured locations and α is the attenuation coefficient owing to the material damping.
𝐶𝑝 = 2𝜋𝑓 (𝑥2−𝑥1
𝜑2−𝜑1) (1)
𝐶𝑔 = 𝑥2−𝑥1
𝑡2−𝑡1 (2)
𝛼 =1
𝑟2−𝑟1𝑙𝑛 (
𝐴2√𝑟2𝑟1
𝐴1) (3)
The wave scattering by the impact damage was measured on the small cross-ply
plates, where the plates had barely visible impact damage from the impact test. An area
of 40 × 40 mm around the impact damage was monitored using a raster scan with step
size of 1 mm. The time traces of the received signals were collected and further processed
in MATLAB. The Hilbert transform was applied to the received signals and the
amplitudes of the signal envelope were used to plot the wave field. In order to understand
the behaviour of the propagating wave field in the composite plate with impact damage,
a visualisation based on the arrival time of each monitored signal was constructed.
Finite Element Model
Guided wave propagation problems were modelled using the finite element (FE) method.
The commercial software package ABAQUS/Explicit was used to simulate the wave
propagation in composite plates. Wave propagation is introduced when the initial
equilibrium is disturbed by the application of forces or displacement on nodes. The
displacement in elements, which can be obtained by integrating the accelerations twice,
is then used to solve the wave propagation problem. The accelerations at the beginning
Page 5
Propagation and scattering of guided waves in composite plates with defects
3732
of the time step are completely determined by the mass and force acting on the elements.
With the explicit method, the state of the elements is advanced through an increment of
time using known values from the previous time step. For computational stability, the
time increment must be smaller than the critical time step. In wave propagation modelling,
the critical time step (Δtcr) can be defined as the transit time at the highest wave speed
through the smallest element in the model. Eq. (4) is used to satisfy the stability
requirement, where le and Cl are respectively the smallest element length and the fastest
wave speed of the material. The length of the element (le) is typically calculated from the
shortest wavelength (λmin) to be analysed, as shown in Eq. (5). For the accuracy of the
simulation, typically at least 10 elements per shortest wavelength are defined [26]. The
chosen time increment must be below the stability limit. These criteria lead to a large
computational memory demand.
Similar to the experimental specimens, three types of composite plates were
modelled. Figure 2 shows the illustration of the plate model, but all plates were modelled
with a large size of 1 × 1 m in order to reduce any unwanted edge reflections and to allow
simpler analysis of the main guided wave properties. Element size of 1 mm in the x- and
y- directions (along the plate) and 0.25 mm in the z-direction (one element per layer
through thickness) was employed, resulting in 8 million elements to model the plate. The
element type was chosen as an eight-node linear brick with reduced integration (C3D8R).
The employed largest element size (1 mm) and time step (0.01 μs) fulfilled the stability
criteria. For the generation of solid homogenous and layered models, the orthotropic
homogenised [0°/90°] and the orthotropic unidirectional [0°] properties were assigned.
The properties were obtained from research by Neau et al. [27]. Rayleigh damping was
set to ß = 30 ns to match the guided wave attenuation measured for the undamaged part
of the composite specimens. Out-of-plane excitation was introduced to generate an
A0 Lamb wave propagating along the plate. The excitation signal consisted of a five-cycle
sinusoidal tone burst modulated by a Hanning window. The pulse was generated with
arbitrary low amplitude based on Eq. (6), as used in the experiments. The excitation
location was placed 100 mm from the centre of the delamination to match the
experimental setup (200 mm for the large delamination model). The out-of-plane
displacement was monitored at the same locations as for the line and circular scans
performed experimentally. A Hilbert transform was used to extract the maximum of the
signal envelopes for each monitoring node.
∆𝑡 ≤ ∆𝑡𝑐𝑟 = 𝑙𝑒
𝐶𝑙 (4)
𝑙𝑒 =𝜆𝑚𝑖𝑛
10 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 (5)
𝐴𝑚𝑝 = 0.5 ∗ (1 − 𝑐𝑜𝑠 (2𝜋𝑓∗𝑡
𝑛𝑜.𝑜𝑓 𝑐𝑦𝑐𝑙𝑒𝑠 )) ∗ (sin (2𝜋𝑓 ∗ 𝑡 ) (6)
To validate both simulation and experimental results, DISPERSE, a software
package developed at Imperial College London by Cawley et al [26], was used to
theoretically predict the guided wave propagation characteristics. Two DISPERSE
models were used to define the composite plates: homogenised and layered models,
similar to those defined in the FE models. Similarly to the FE simulations, each model
was defined using lossy orthotropic stiffness properties [27] where both real and
imaginary stiffness properties were used as the inputs for the material properties. From
Neau et al [27], the uncertainty of the imaginary properties is significantly larger than for
Page 6
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3733
the real constants. Thus, it is expected that the attenuation measurements are going to
have larger errors than the velocity measurements.
Figure 2. Illustration of the plate geometry; the excitation and monitoring points are
placed in the middle of plate for the generation and detection of the A0 mode; corner
nodes number identified as nb_i (i = 1, 2, 3…n); edges 1, 2, 3 and 4 were set as the
boundaries of the plate.
RESULTS AND DISCUSSION
The propagation of the A0 mode was measured using a non-contact laser interferometer
and compared to the results of the Finite Element (FE) simulations as well as to the
DISPERSE semi-analytical predictions. The aim of this study is to investigate the wave
dispersion and attenuation of the A0 mode in anisotropic plates.
Group and Phase Velocities
This section reports the velocity dispersion characteristic of the A0 guided wave mode in
three types of undamaged composite plates, focused on the measurements in the 0°
direction. Figure 3 shows the comparisons of the measured group and phase velocities
together with predictions by DISPERSE and FE analysis. Comparable results can be
observed for the DISPERSE and FE predictions (within 1% error) in calculating the group
and phase velocities for three different types of plates. This is as expected because both
simulations were using the same stiffness coefficients (real part), which determines the
guided wave velocity in plates. Comparing the predicted results to the experimental
results, the measured phase velocities are scattered around the predicted values
reasonably well. A larger error can be observed at the higher frequency region with the
largest (approximately less than 30% error) obtained from the 2 mm cross-ply plate. In
term of velocity dispersion, it can be seen that the A0 mode is highly dispersive in the
frequency range below 50 kHz for all composite plates. This wave dispersion is
undesirable in inspection systems because there will be an increase in the pulse width and
a decrease in the amplitude with propagation distance owing to the broadening
Page 7
Propagation and scattering of guided waves in composite plates with defects
3734
distribution of the wave energy [21]. The reduction in amplitude limits the propagation
distance, and the increase in signal duration worsens the resolution that can be obtained.
Figure 3. Measured and predicted velocities for the A0 mode propagation in the
composite plates; a) and (b) for the 3.6 mm cross-ply plate, (c) and (d) for the 2 mm
cross-ply plate, and (e) and (f) for the 3.6 mm UD plate; 100 kHz; measured in 0°
direction.
On the contrary, Figure 3 also shows that the dispersion of the A0 mode in all
plates is smaller above 50 kHz. Within the frequency range of the excited wave packet
(50 kHz to 150 kHz), there were small dispersions in the group velocities. Wilcox et al.
[21] explained that different frequency components in a wave packet travel at different
speeds, so the shape of the wave packet is expected to have a small distortion while
traveling owing to the differences in the arrival times of each frequency component. This
0.1 0.2 0.3 0.4 0.5 0.60
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Gro
up
ve
locity [m
/s]
Experiment
FEA
Disperse:Homogenized
Disperse:Layered
0.1 0.2 0.3 0.4 0.5 0.60
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Ph
ase
ve
locity [m
/s]
Experiment
FEA
Disperse:Homogenized
Disperse:Layered
0 0.1 0.2 0.3 0.40
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Gro
up
ve
locity [m
/s]
Exp. Plate 1
Exp. Plate 2
FEA
Disperse:Homogenized
Disperse:Layered
0 0.1 0.2 0.3 0.40
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Ph
ase
ve
locity [m
/s]
Exp. Plate 1
Exp. Plate 2
FEA
Disperse:Homogenized
Disperse:Layered
0.1 0.2 0.3 0.4 0.5 0.60
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Gro
up
ve
locity [m
/s]
Experiment
FEA
Disperse:Homogenized
Disperse:Layered
0.1 0.2 0.3 0.4 0.5 0.60
500
1000
1500
2000
Frequency-Thickness [MHz.mm]
Ph
ase
ve
locity [m
/s]
Experiment
FEA
Disperse:Homogenized
Disperse:Layered
Page 8
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3735
means that the wave packet retains its shape as it travels and the differences in the arrival
times are very small. This feature is desirable for the SHM of composites because a
dispersive wave packet will lead to a complicated signal processing [21]. For the defect
characterisation measurements, a 100 kHz frequency was chosen as the nominal
excitation frequency as its wave packet is less dispersive.
Attenuation
Both DISPERSE models of the three types of plates were further analysed to estimate the
attenuation coefficient in the frequency range from 50 kHz to 150 kHz. From Figure 4,
the overall trend of the simulated models is comparable to the experimental results,
although it gives slightly higher values. The attenuation increased with frequency for all
three types of composite plates, as expected for the A0 mode. A small variation of the
measured coefficients at frequencies below 100 kHz can be observed, which could be
owing to measurement errors. At frequencies above 100 kHz the values became more
stable. For the 3.6 mm cross-ply plate (Figure 4), the measured attenuation values
increased from approximately 0.05 dB/mm at 50 kHz to 0.15 dB/mm at 150 kHz. This
gives a rough estimation of an approximately 0.10 dB/mm increment over the frequency
range. Similarly, the attenuation coefficients for the 3.6 mm UD plate were measured as
0.001 dB/mm at 50 kHz and increased to approximately 0.11 dB/mm at 150 kHz, lower
than for the 3.6 mm cross-ply plate. Since the results compare reasonably well, these
attenuation values hold important information to determine how far the A0 mode can
travel at certain excitation frequencies and can be used for future reference. The increase
in attenuation with frequency imposes an upper frequency limit on inspections. From
these attenuation results, the frequency-dependency of the A0 mode wave attenuation has
been shown and validated.
Comparisons between the attenuation coefficients of the three different composite
plates show the influence of plate thickness, material properties and fibre arrangement.
From Figure 4, it can be seen that the 2 mm plate has lower wave attenuation at 100 kHz
(α = 0.046 dB/mm) than the 3.6 mm plate (α = 0.099 dB/mm). Referring to literature for
a qualitative comparison, Herrmann et al. [28, 29] also showed a similar behaviour of the
attenuation coefficient for the A0 mode in 12-ply and 16-ply UD plates. The attenuation
values were respectively = 0.025 dB/mm and 0.06 dB/mm (in the 0° direction). Two
reasons seem to correlate to this behaviour: (i) the material thickness and (ii) differences
in the materials used for their fibre-matrix system. Thicker materials generally show
greater damping owing to increased energy absorption. Various studies have shown that
the wave absorption coefficient of a composite material is a function of thickness and
porosity. Furthermore, Prosser [30] discussed the influence of plate thickness on wave
attenuation. It was demonstrated that the thicker plate increased the wave attenuation. To
relate to the second reason, Biwa [31] indicated that a major governing factor of wave
attenuation is the viscoelastic absorption in the matrix. This results in an increase of the
wave attenuation with increasing matrix content. These seem to support the findings
presented here, where the 3.6 mm cross-ply plate has higher matrix content (40%) than
the 2 mm cross-ply plate (35%). However, it should be noted that both types of plates
were made of different fibre-matrix systems, which could additionally contribute to the
different wave attenuation of both plates.
From Figure 4, a comparison between the 3.6 mm UD and the 3.6 mm cross-ply
plates shows that the fibre arrangement (alternately arranged into the 0° and 90°
directions) contributed to higher attenuation than the one with all fibres arranged in the
0° direction. Although both plates are made of the same number of plies and have the
Page 9
Propagation and scattering of guided waves in composite plates with defects
3736
same material properties, the wave attenuation differed by 13%. This could be owing to
the direction of measurements taken along the fibre direction (0°) of the UD plate, in
which it has higher stiffness properties than in the cross-ply plate. In comparison to what
has been published by other researchers, Ono and Gallego [32] measured attenuation
coefficient (α) of the A0 mode = 0.08 dB/mm in a 16 cross-ply composite plate and α =
0.06 dB/mm in a 16-ply UD plate in the 0° direction. Although the measurements were
performed at 300 kHz, their results are qualitatively similar to the ones presented here,
where the attenuation in the UD plate is lower than the cross-ply plate, although their
number of plies and materials were the same.
(a) (b)
(c)
Figure 4. Measured frequency-dependent attenuation coefficient together with
DISPERSE predictions; corrected for geometrical beam spreading; excitation frequency
50–150 kHz; measured in 0° direction; (a) 24-cross ply; (b) 8-cross ply; (c) 24-
unidirectional composite plates.
Scattering of Guided Waves at Impact Damage
Figure 5 presents the guided wave fields at various time snapshots for damaged composite
plates 1 and 2. It can be seen that the incident wave interacts with the impact damage and
causes scattering within the damaged region. Relatively weak scattering by the damage
is present when the incident wave arrives (Figure 5a and Figure 5c), then a significant
increase of the scattering is apparent when the wave has travelled past the damaged area
(Figure 5b and Figure 5d). Based on the observation, a significant portion of the waves is
20 40 60 80 100 120 140 1600
0.05
0.1
0.15
0.2
0.25
Frequency [kHz]
Atte
nu
atio
n [d
B/m
m]
Exp.
Disperse: Homogenized
Disperse: Layered
20 40 60 80 100 120 140 1600
0.05
0.1
0.15
0.2
0.25
Frequency [kHz]
Atte
nu
atio
n [d
B/m
m]
Exp. Plate 1
Exp. Plate 2
Disperse: Homogenized
Disperse: Layered
20 40 60 80 100 120 140 1600
0.05
0.1
0.15
0.2
0.25
Frequency [kHz]
Atte
nu
atio
n [d
B/m
m]
Exp.
Disperse: Homogenized
Disperse: Layered
Page 10
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3737
also reflected back from the exit of the impact damage. These reflected waves travel
within the damaged region and keep reflecting back at both the entrance and exit of the
impact damage. As a consequence of these multiple reflections, a considerable amount of
guided wave energy is trapped inside the impact damage area. This causes an increase in
the amplitude of the wave in that region. The amplitude of the transmitted waves
diminished noticeably after passing through the impacted area. Similar wave behaviour
was also observed by Sohn et al. [33], where the measurement was performed on a quasi-
isotropic composite plate. In the FE results in a different publication [34], a similar
reduction in the amplitude past the delamination area was observed, which is in agreement
with this experimental measurement. Comparing both specimens, which were treated
with the same 7.4 J impact energy, it can be seen that guided wave scattering in both
plates is unique and such variations are expected owing to the complexity in the failure
mechanism of impact damage.
Figure 5. Experimental guided wave displacement fields in two composite plates. Plate
No. 1, snapshot time: (a) 20 μs and (b) 30 μs. Plate No. 2, snapshot time: (c) 20 μs and
(d) 30 μs. 7.4 J impact; 40 × 40 mm scanned area.
Figure 6 presents an image of the maximum amplitude of the enveloped signal
over the damaged areas in both specimens. Areas of higher amplitudes can be seen that
occur close to the impact location centre (x = 20 mm, y = 20 mm). This indicates the
presence of severe damage, such as delamination or fibre and matrix cracking, and
matches reasonably well with the visually observed size of the impact damage on the
plates. Meanwhile, the undamaged area is represented by the low amplitude distribution.
From the figure, three different zones can be observed and each zone has its own wave
(a) (b)
(c) (d)
Incident wave
Incident wave
x = 40 mm
y = 40 mm
y = 40 mm
x = 40 mm
Page 11
Propagation and scattering of guided waves in composite plates with defects
3738
propagation behaviour; (i) zone 1: before the impact damage (x < 15 mm), (ii) zone 2:
across the impact damage (15 mm < x < 25 mm) and (iii) zone 3: behind the impact
damage (x > 25 mm). The first zone (before damage) shows the incident waves
propagating towards the impact damage location.
Figure 6: Experimental guided wave pulses across damaged area of composite plate: a)
plate 1; b) plate 2. Frequency 100 kHz; 40 × 40 mm area with impact location at centre
(x = 20 mm, y = 20 mm).
Reflected waves propagating back towards the excitation source can also be seen
there. Some periodical increase and decrease of the amplitude is visible in the region
around x = 15 mm, potentially indicating the interference between the incident and the
reflected waves at the impact damage. In zone 2, high amplitudes of the A0 mode signals
are visible, which could indicate the multiple reflections and scattering events within the
impact damage area. Meanwhile, in zone 3, the transmitted waves propagating out from
the damaged area are seen to be blocked in certain direction with significantly reduced
10 20 30 40
5
10
15
20
25
30
35
40
X-position [mm]
Y-p
ositio
n [m
m]
0.5
1
1.5
2
10 20 30 40
5
10
15
20
25
30
35
40
X-position [mm]
Y-p
ositio
n [m
m]
0.5
1
1.5
2
(b)
(a)
Incident wave
Incident wave
Page 12
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3739
signal amplitudes. Considerable waves transmitted across the impact damage indicate that
probably little mode conversion occurred. The cross-section of the impact damage in plate
no. 1 can be roughly approximated with a length of 10 mm (x-axis) and a width of 20 mm
(y-axis), identified by the higher amplitude maxima. For plate no. 2, the cross section area
of the impact damage can be estimated to be about 10 × 25 mm. Comparing to the FE
simulation results [34], good agreement between the experimental and simulated results
was achieved. Similar wave propagation behaviour before and within the impact damage
can be observed from both results. The increase in amplitudes identifies the localisation
of the impact damage in the measured area.
CONCLUSIONS
The use of the A0 guided ultrasonic wave mode excited at 100 kHz for the detection and
characterisation of impact damage in composite plates has been shown, with a view to
employing this methodology for composite structural health monitoring. The measured
wave velocities and attenuation are reasonably in agreement with the Finite Element
simulated results. The highly dispersive behaviour of the A0 can be observed at
frequencies below 50 kHz. Meanwhile, a repeatable scattering pattern at the impact
damage was observed experimentally. The location of the impact damage can be
identified by the increase in amplitude and subsequent significant amplitude reduction
past the damage location. Good agreement between the experimental and finite element
results was obtained. This study demonstrated that low frequency A0 guided wave mode
generated by a piezoelectric transducer can be successfully employed to monitor impact
damage in composite plates. The potential of guided waves for monitoring composite
structures has been shown in this study, and a better understanding of the guided wave
interaction with defects was achieved.
ACKNOWLEDGEMENTS
The author gratefully acknowledges the support in funding by Universiti Teknologi
MARA (UiTM), Malaysia, especially from the Geran Dana Pembudayaan Penyelidikan
(600-RMI/RAGS 5/3 (12/2015).
REFERENCES
[1] Richardson MOW, Wisheart MJ. Review of low-velocity impact properties of
composite materials. Composites Part A: Applied Science and Manufacturing.
1996;27:1123-31.
[2] Perillo G, Vedivik NP, Echtermeyer AT. Damage development in stitch bonded
GFRP composite plates under low velocity impact: Experimental and numerical
results. Journal of Composite Materials. 2014;49:601-15.
[3] Shyr T-W, Pan Y-H. Impact resistance and damage characteristics of composite
laminates. Composite Structures. 2003;62:193-203.
[4] Topac OT, Gozluklu B, Gurses E, Coker D. Experimental and computational
study of the damage process in CFRP composite beams under low-velocity
impact. Composites Part A: Applied Science and Manufacturing. 2017;92:167-
82.
Page 13
Propagation and scattering of guided waves in composite plates with defects
3740
[5] Choi HY, Wu H-YT, Chang F-K. A New Approach toward understanding damage
mechanisms and mechanics of laminated composites due to low-velocity impact:
Part II-analysis. Journal of Composite Materials. 1991;25:1012-38.
[6] Lou X, Cai H, Yu P, Jiao F, Han X. Failure analysis of composite laminate under
low-velocity impact based on micromechanics of failure. Composite Structures.
2017;163:238-47.
[7] Huzni S, Fonna S, Arifin A. Finite element modeling of delamination process on
composite laminate using cohesive elements. International Journal of Automotive
and Mechanical Engineering. 2013;7:1023-30.
[8] Wisnom MR. The role of delamination in failure of fibre-reinforced composites.
Philosophical Transactions of the Royal Society A: Mathematical, Physical and
Engineering Sciences. 2012;370:1850-70.
[9] Papa I, Lopresto V, Simeoli G, Langella A, Russo P. Ultrasonic damage
investigation on woven jute/poly (lactic acid) composites subjected to low
velocity impact. Composites Part B: Engineering.
[10] Hafizi ZM, Epaarachchi J, Lau KT. An investigation of acoustic emission signal
attenuation for monitoring of progressive failure in fiberglass reinforced
composite laminates. International Journal of Automotive and Mechanical
Engineering. 2013;8:1442-56.
[11] Endrizzi M, Murat BIS, Fromme P, Olivo A. Edge-illumination X-ray dark-field
imaging for visualising defects in composite structures. Composite Structures.
2015;134:895-9.
[12] Liang T, Ren W, Tian GY, Elradi M, Gao Y. Low energy impact damage detection
in CFRP using eddy current pulsed thermography. Composite Structures.
2016;143:352-61.
[13] P. Fromme, P. D. Wilcox, M. J. S. Lowe, P. Cawley. On the development and
testing of a guided ultrasonic wave array for structural integrity monitoring. IEEE
Transactions on Ultrasonics, Ferroelectrics and Frequency Control. 2006;53:8.
[14] Rose J L. Ultrasonic guided waves in structural health monitoring. Key
Engineering Material. 2004;273-275:7.
[15] Fromme P, Sayir MB. Measurement of the scattering of a Lamb wave by a through
hole in a plate. The Journal of the Acoustical Society of America. 2002;111:1165-
70.
[16] Ishak SI, Liu GR, Lim SP, Shang HM. Characterization of delamination in beams
using fexural wve sattering aalysis. Journal of Vibration and Acoustics.
2001;123:421-7.
[17] Ju TH, Datta SK. Scattering of Impact Wave by a Crack in Composite Plate. In:
Thompson DO, Chimenti DE, editors. Review of Progress in Quantitative
Nondestructive Evaluation: Volume 10B. Boston, MA: Springer US; 1991. p.
1515-22.
[18] Lowe MJS, Neau G, Deschamps M. Properties of Guided Waves in Composite
Plates, and Implications for NDE. AIP Conference Proceedings. 2004;700:214-
21.
[19] Rose JL. A Baseline and Vision of Ultrasonic Guided Wave Inspection Potential.
Journal of Pressure Vessel Technology. 2002;124:273-82.
[20] Wilcox PD, Lowe MJS, Cawley P. Mode and Transducer Selection for Long
Range Lamb Wave Inspection. Journal of Intelligent Material Systems and
Structures. 2001;12:553-65.
Page 14
Murat and Fromme / International Journal of Automotive and Mechanical Engineering 13(3) 2016 3728-3741
3741
[21] Wilcox P, Lowe M, Cawley P. The effect of dispersion on long-range inspection
using ultrasonic guided waves. NDT & E International. 2001;34:1-9.
[22] Su Z, Ye L, Lu Y. Guided Lamb waves for identification of damage in composite
structures: A review. Journal of Sound and Vibration. 2006;295:753-80.
[23] Lu Y, Ye L, Su Z, Yang C. Quantitative assessment of through-thickness crack
size based on Lamb wave scattering in aluminium plates. NDT & E International.
2008;41:59-68.
[24] Karunasena WM, Shah AH, Datta SK. Plane strain wave scattering by cracks in
laminated composite plates. Journal of Engineering Mechanics. 1991;117:1738-
54.
[25] Ng CT, Veidt M. Scattering analysis of fundamental anti-symmetric Lamb wave
at delaminations in composite laminates. Australian Journal of Mechanical
Engineering. 2011;8:197-205.
[26] Alleyne DN, Cawley P. Optimization of lamb wave inspection techniques. NDT
& E International. 1992;25:11-22.
[27] Neau G, Lowe MJS, Deschamps M. Propagation of lamb waves in anisotropic and
absorbing plates: Theoretical derivation and experiments. AIP Conference
Proceedings. 2002;615:1062-9.
[28] Herrmann F, Jochim B, Oßwald P, Cai L, Pitsch H, Kohse-Höinghaus K.
Experimental and numerical low-temperature oxidation study of ethanol and
dimethyl ether. Combustion and Flame. 2014;161:384-97.
[29] Schubert KJ, Herrmann AS. On attenuation and measurement of Lamb waves in
viscoelastic composites. Composite Structures. 2011;94:177-85.
[30] Prosser WH, Seale MD, Smith BT. Time-frequency analysis of the dispersion of
Lamb modes. The Journal of the Acoustical Society of America. 1999;105:2669-
76.
[31] Biwa S, Watanabe Y, Ohno N. Analysis of wave attenuation in unidirectional
viscoelastic composites by a differential scheme. Composite Science and
Technology. 2003;63:237-47.
[32] Ono K, Gallego A. Attenuation of lamb waves in CFRP plates. Journal of
Acoustic Emission. 2012;30:109-23.
[33] Sohn H, Dutta D, Yang JY, Park HJ, DeSimio M, Olson S, et al. Delamination
detection in composites through guided wave field image processing. Composite
Science and Technology. 2011;71:1250-6.
[34] Murat BIS, Khalili P, Fromme P. Scattering of guided waves at delaminations in
composite plates. The Journal of the Acoustical Society of America.
2016;139:3044-52.